Storing and Accessing Live Mashup Content in the Cloud Krzysztof Ostrowski, Ken Birman Cornell University

Agenda

1.A new, versatile storage abstraction: Checkpointed Channel (CC) 2.A new web application architecture: A web of (hyperlinked) CCs Agenda

Introduction

Storing Content in the Cloud server in a data center client-server interactions cloud content edge

Storing Content in the Cloud server in a data center client-server interactions cloud content edge

server in a data center client-server interactions cloud Storing Content in the Cloud content edge

server in a data center client-server interactions cloud Storing Content in the Cloud content edge

server in a data center client-server interactions cloud Storing Content in the Cloud content edge

server in a data center client-server interactions cloud Storing Content in the Cloud content edge

server in a data center client-server interactions cloud Storing Content in the Cloud content edge

Storing Content in the Cloud content client-server interactions server in a data center update relayed through the server

server in a data center Storing Content in the Cloud explicit request, e.g. HTTP GET content cloud edge

server in a data center Storing Content at the Edge edge content replica content replica content replica peer-to-peer interactions

server in a data center Storing Content at the Edge edge peer-to-peer interactions content replica content replica content replica

server in a data center Storing Content at the Edge edge peer-to-peer interactions content replica content replica content replica

server in a data center Storing Content at the Edge edge peer-to-peer interactions content replica content replica content replica

server in a data center Storing Content at the Edge edge peer-to-peer interactions content replica content replica content replica updated

server in a data center Storing Content at the Edge edge peer-to-peer interactions content replica content replica content replica updated multicast

server in a data center Storing Content at the Edge edge content replica content replica content replica updated

server in a data center Storing Content at the Edge edge content replica content replica content replica replicated state machine

server in a data center Storing Content at the Edge edge content replica content replica content replica manage membership

server in a data center Storing Content at the Edge edge data control

server in a data center Storing Content at the Edge edge data control

Cloud vs. Edge? server in a data center edge peer-to-peer interactions content replica content replica content replica server in a data center client- server interactions cloud content edge

server in a data center Cloud vs. Edge? edge

server in a data center Cloud vs. Edge? edge

server in a data center Cloud vs. Edge? edge

server in a data center Cloud vs. Edge? edge no replicas left

Cloud vs. Edge? server in a data center client-server interactions cloud content edge flow of updates bottleneck

Cloud vs. Edge? server in a data center client-server interactions cloud content edge flow of updates bottleneck

Cloud vs. Edge? server in a data center client-server interactions cloud content edge flow of updates bottleneck

Cloud vs. Edge? server in a data center client-server interactions cloud content edge flow of updates bottleneck

Cloud vs. Edge? server in a data center client-server interactions cloud content edge flow of updates bottleneck

Cloud vs. Edge? Server capacity: SecondLife: ~40 clients/server Second Life and the New Generation of Virtual Worlds. S. Kumar et al IEEE Computer, 41(9):46-53

Cloud vs. Edge? Server capacity: SecondLife: ~40 clients/server virtual island

Cloud Edge? server in a data center edge peer-to-peer interactions content replica content replica content replica server in a data center client- server interactions cloud content edge persistencescalability + and

Extending Cloud to the Edge server edge cloud

Extending Cloud to the Edge cloud server edge

peer-to-peer interactions client-server interactions Extending Cloud to the Edge cloud server content

peer-to-peer interactions client-server interactions Extending Cloud to the Edge cloud server content distributed protocol (client-server, peer-to-peer, etc.) content

Extending Cloud to the Edge cloud server edge

There are: 40+ times more PCs than servers (billions today, to double in five years) client PCs per Google server 40,000+ times more client PCs purchased each year by home users than new servers purchased each year by Microsoft => a lot of computational power and resources at the edge that are greatly underutilized Edge is Larger than the Cloud

There are: 40+ times more PCs than servers (billions today, to double in five years) client PCs per Google server 40,000+ times more client PCs purchased each year by home users than new servers purchased each year by Microsoft => a lot of computational power and resources at the edge that are greatly underutilized Edge is Larger than the Cloud

There are: 40+ times more PCs than servers (billions today, to double in five years) client PCs per Google server 40,000+ times more client PCs purchased each year by home users than new servers purchased each year by Microsoft => a lot of computational power and resources at the edge that are greatly underutilized Edge is Larger than the Cloud

There are: 40+ times more PCs than servers (billions today, to double in five years) client PCs per Google server 40,000+ times more client PCs purchased each year by home users than new servers purchased each year by Microsoft => a lot of computational power and resources at the edge that are greatly underutilized Edge is Larger than the Cloud

There are: 40+ times more PCs than servers (billions today, to double in five years) client PCs per Google server 40,000+ times more client PCs purchased each year by home users than new servers purchased each year by Microsoft => a lot of computational power and resources at the edge that are greatly underutilized Edge is Larger than the Cloud

Unify different content access models: Centralized vs. replicated Centralized: web service, database Replicated: distributed P2P replication protocols Persistent vs. temporary Persistent: server in a data center Temporary: collaboration session formed ad hoc Server-side vs. client-side Shared public vs. shared private Public: web service or collaboration session Private: session state, cookies, etc. Stateful vs. stateless Stateful: variables, files, objects, database tables Stateless: video or notification streams New Storage Abstraction

Unify different content access models: Centralized vs. replicated Centralized: web service, database Replicated: distributed P2P replication protocols Persistent vs. temporary Persistent: server in a data center Temporary: collaboration session formed ad hoc Server-side vs. client-side Shared public vs. shared private Public: web service or collaboration session Private: session state, cookies, etc. Stateful vs. stateless Stateful: variables, files, objects, database tables Stateless: video or notification streams New Storage Abstraction

Unify different content access models: Centralized vs. replicated Centralized: web service, database Replicated: distributed P2P replication protocols Persistent vs. temporary Persistent: server in a data center Temporary: collaboration session formed ad hoc Server-side vs. client-side Shared public vs. shared private Public: web service or collaboration session Private: session state, cookies, etc. Stateful vs. stateless Stateful: variables, files, objects, database tables Stateless: video or notification streams New Storage Abstraction

Unify different content access models: Centralized vs. replicated Centralized: web service, database Replicated: distributed P2P replication protocols Persistent vs. temporary Persistent: server in a data center Temporary: collaboration session formed ad hoc Server-side vs. client-side Shared public vs. shared private Public: web service or collaboration session Private: session state, cookies, etc. Stateful vs. stateless Stateful: variables, files, objects, database tables Stateless: video or notification streams New Storage Abstraction

Unify different content access models: Centralized vs. replicated Centralized: web service, database Replicated: distributed P2P replication protocols Persistent vs. temporary Persistent: server in a data center Temporary: collaboration session formed ad hoc Server-side vs. client-side Shared public vs. shared private Public: web service or collaboration session Private: session state, cookies, etc. Stateful vs. stateless Stateful: variables, files, objects, database tables Stateless: video or notification streams New Storage Abstraction

Unify different content access models: Centralized vs. replicated Centralized: web service, database Replicated: distributed P2P replication protocols Persistent vs. temporary Persistent: server in a data center Temporary: collaboration session formed ad hoc Server-side vs. client-side Shared public vs. shared private Public: web service or collaboration session Private: session state, cookies, etc. Stateful vs. stateless Stateful: variables, files, objects, database tables Stateless: video or notification streams New Storage Abstraction

Checkpointed Channels (CC)

Architecture

network messages

the edge may “cut through” the channel content physically resides outside the channel… …but it could also be cached in it

the edge may “cut through” the channel content physically resides outside the channel… …but it could also be cached in it

content physically resides outside the channel… …but it could also be cached in it the edge may “cut through” the channel

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Dynamics (local) Checkpointed Communication Channel (CC) U C UU U initial checkpoint

Dynamics (local) Checkpointed Communication Channel (CC) C initial checkpointmultiple updates UUU U

Dynamics (local) Checkpointed Communication Channel (CC) C initial checkpointmultiple updates UUU U

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Dynamics (local) Checkpointed Communication Channel (CC) U updates UU

Dynamics (local) Checkpointed Communication Channel (CC) U updates UU

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Interfaces Checkpointed Communication Channel (CC) a writable stream of checkpoints and updates

Dynamics (local) Checkpointed Communication Channel (CC) request checkpoint R

Dynamics (local) Checkpointed Communication Channel (CC) request checkpoint R

Dynamics (local) Checkpointed Communication Channel (CC) C checkpoint

Dynamics (local) Checkpointed Communication Channel (CC) C checkpoint

Dynamics (global) Checkpointed Channel (CC)

Dynamics (global) Checkpointed Channel (CC) connect

Dynamics (global) Checkpointed Channel (CC) connect initialize

Dynamics (global) connect initialize need checkpoint

Dynamics (global) initialize need checkpoint R

Dynamics (global) request checkpoint need checkpoint

Dynamics (global) request checkpoint R

Dynamics (global) initialize request checkpoint need checkpoint R

Dynamics (global) request checkpoint need checkpoint provide checkpoint

Dynamics (global) request checkpoint provide checkpoint C

Dynamics (global) provide checkpoint C

Dynamics (global) C transfer checkpoint

Dynamics (global) transfer checkpoint C

Dynamics (global) C checkpoint transfer checkpoint

Dynamics (global) C checkpoint deliver checkpoint

Dynamics (global) deliver checkpoint C

Dynamics (global) consume checkpoint C deliver checkpoint

Dynamics (global) consume checkpoint C wants to update

Dynamics (global) consume checkpoint C wants to update U

Dynamics (global) wants to update U request update

Dynamics (global) update request update U

Dynamics (global) U propagate update

Dynamics (global) propagate update UUU

Dynamics (global) propagate update U UU U

Dynamics (global) U UU update U deliver update

Dynamics (global) U UU U deliver update

Dynamics (global) deliver update U UU U

Dynamics (global) U UU U consume update

Semantics UCC'C' + = earlier checkpoint or state new checkpoint or state incremental update

Semantics UCC'C' + = earlier checkpoint or state new checkpoint or state incremental update

Semantics UCC'C' + = earlier checkpoint or state new checkpoint or state incremental update

Semantics CkCk checkpoint multiple updates U1U1 U2U2 UnUn

Semantics CkCk U1U1 U2U2 UnUn

CkCk U1U1 U2U2 UnUn C k+1 += CkCk U1U1

Semantics CkCk U1U1 U2U2 UnUn C k+1 C k+2 += C k+1 U2U2

Semantics CkCk U1U1 U2U2 UnUn C k+1 C k+2 C k+n + = C k+n-1 UnUn

Semantics CkCk U1U1 U2U2 UnUn C k+1 C k+2 C k+n

Semantics CkCk U1U1 U2U2 UnUn C k+1 C k+2 C k+n

Semantics C1C1 U2U2 U3U3 U5U5 C4C4 C6C6

C1C1 U2U2 U3U3 U5U5 C4C4 C6C6 C2C2 C3C3 C5C5

C1C1 U2U2 U3U3 U5U5 C4C4 C6C6 C2C2 C3C3 C5C5

C1C1 U2U2 U3U3 U5U5 C4C4 C6C6 C2C2 C3C3 C5C5

C1C1 CnCn C2C2 C3C3 “master sequence”

Semantics C1C1 CnCn C2C2 C3C3 “master sequence” C1C1 U n-1 U1U1 U2U2 submitted updates

Semantics C1C1 CnCn C2C2 C3C3 “master sequence” C1C1 C5C5 C8C8 C9C9 C 15 every application observes a subset of the master sequence

Semantics C1C1 CnCn C2C2 C3C3 “master sequence” C1C1 C5C5 C8C8 C9C9 C 15 every application observes a subset of the master sequence …and nobody lags behind

Types Checkpointed Communication Channel (CC) can be classified based on: the type of checkpoints, the type of updates, (and many other factors we won’t dicuss)

Types Checkpointed Communication Channel (CC) can be classified based on: the type of checkpoints, the type of updates, (and many other factors we won’t dicuss)

Types C U XML documents standardized edits on XML documents XML Channel (XC)

Types C U XML documents standardized edits on XML documents XML Channel (XC)

Types C U standardized edits on XML documents XML Channel (XC) XML documents

Types C U XML Channel (XC) XML documents standardized edits on XML documents

Types Checkpointed Channel (CC) Text ChannelBinary Channel XML Channel (XC) RSS Channel XHTML Channel XAML Channel Reference Channel Collection

Types Checkpointed Channel (CC) Text ChannelBinary Channel XML Channel (XC) RSS Channel XHTML Channel XAML Channel Reference Channel Collection

Types Checkpointed Channel (CC) Text ChannelBinary Channel XML Channel (XC) RSS Channel XHTML Channel XAML Channel Reference Channel Collection

Types Checkpointed Channel (CC) Text ChannelBinary Channel XML Channel (XC) RSS Channel XHTML Channel XAML Channel Reference Channel Collection

Live Distributed Objects (LO)

Ordinary Web Applications displayUI components

Ordinary Web Applications displayUI components

Ordinary Web Applications displayUI components

Ordinary Web Applications displayUI components

Ordinary Web Applications displayUI components

Ordinary Web Applications displayUI components

Ordinary Web Applications side by side nested

Ordinary Web Applications

Live Objects Applications side by side nested front-end to back-end

Live Objects Applications side by side nested front-end to back-end data pipeline

Live Objects Applications side by side nested front-end to back-end data pipeline UI component

Live Objects Applications side by side nested front-end to back-end data pipeline UI component MPEG-2 decoder

Live Objects Applications side by side nested front-end to back-end data pipeline UI component MPEG-2 decoder AES decryptor

Live Objects Applications side by side nested front-end to back-end data pipeline UI component MPEG-2 decoder AES decryptor reliable transport

Live Objects Applications side by side nested front-end to back-end data pipeline UI component MPEG-2 decoder AES decryptor reliable transport unreliable transport

… side by side nested front-end to back-end data pipeline

… side by side nested front-end to back-end

… side by side nested front-end to back-end

… side by side nested front-end to back-end

… side by side nested front-end to back-end

… side by side nested front-end to back-end

Live Objects Applications UI component MPEG-2 decoder AES decryptor reliable multicast IP multicast recovery protocol hole-punching UDP transport TCP transport membership client web service proxy

Live Objects Applications UI component MPEG-2 decoder AES decryptor reliable multicast IP multicast recovery protocol hole-punching UDP transport TCP transport membership client web service proxy

Live Objects Applications UI component MPEG-2 decoder AES decryptor reliable multicast IP multicast recovery protocol hole-punching UDP transport TCP transport membership client web service proxy

Live Objects Applications UI component MPEG-2 decoder AES decryptor reliable multicast IP multicast recovery protocol hole-punching UDP transport TCP transport membership client web service proxy

Live Objects Applications UI component MPEG-2 decoder AES decryptor reliable multicast IP multicast recovery protocol hole-punching UDP transport TCP transport membership client web service proxy

Live Objects Applications data “pipeline”

Live Objects Applications data pipeline data pipeline

Live Objects Applications data pipeline data pipeline data pipeline

Live Objects Applications XML

References XML activate channel reference channel proxy

References XML activate channel reference channel proxy

References XML activate channel reference channel proxy

References XML channel reference channel proxy points to

References XML channel reference channel proxy points to channel content XML

References XML points to XML

References XML points to XML

References XML

Ordinary Web Applications side by side nested XML markup

Live Objects Applications XML

Live Objects Applications

space

Live Objects Applications space plane 1 XML

Live Objects Applications space plane 1 plane 2 XML

Live Objects Applications space plane 1 plane 2 building 1 XML

Live Objects Applications space plane 1 plane 2 building 1 map XML

Live Objects Applications tile 1 XML

Live Objects Applications tile 1 XML tile 2 XML

Live Objects Applications

desktop

Live Objects Applications desktop text 2 XML

Live Objects Applications desktop text 2 XML image 1 XML

Live Objects Applications desktop text 2 XML image 1 XML desktop 2 XML

Live Objects Applications desktop text 2 XML image 1 XML desktop 2 XML text 3 XML

Live Objects Applications desktop 1

Live Objects Applications desktop 1 XML desktop 2

Live Objects Applications desktop 1 XML desktop 2 desktop 3 XML

Live Objects Applications desktop 1 XML desktop 2 desktop 3 XML text 1 XML

Live Objects Applications desktop 1 XML desktop 2 desktop 3 XML text 1 XML

Conclusions

1.Cloud and edge technologies can work together… …but we need a new storage abstraction. 2.Checkpointed Channels (CCs) a)Support a variety of protocols: client-server, P2P, distributed replication, append log files, etc., b)Can be used at any level of the protocol stack, at the fronted and at the backend, c)Can efficiently store structured mashup content, d)Can be hyperlinked into webs of CCs; these could serve as a basis for a new Web architecture. Conclusions

Thanks